Study and Develop a System for Automated Bird Count and Data Collection

The objective of this project is to design a system that could automatically calculate the number of chicken and their varying size.

The system consists of five parts. The first part is to study the deployment environment. The second is the sensor selection and hardware design. The third is the code algorithm and method to detect the presence of chicken. The fourth illustrates the communication interface. Lastly, designing of graphic user interface will be discussed.

The deployment environment is very humid caused by water condensation. Wireless deployment becomes unstable due to the varying water condensation intensity. If a cable is hung for testing, water droplets will be observed dripping from the cable within 3 minutes. The place will be washed using water pipe, so water resistance and vibration will be a concern to the design too.

Figure 1. Table show the test result of different sensor

After testing various sensors, light intensity was decided upon. It is also more cost effective.

Due to the chickens’ varying size and improper position, the system is registering a false count while detection is from the neck down to the bottom. So, a timer to detect the feet was introduced. The algorithm was able to provide an accurate count of the number chickens and also the hanging position.

Figure 2. Captured Data, green is chicken detection profile, yellow is hanger detection profile
Figure 3 shows how the  detector/sensor works.

When the chicken feet and the hanger are detected, the signal is sent to the LED count display and captured by application.

Designing with wireless communication was decided upon after completion of field test on wireless receiving sensitivity. However, after 3 weeks, when the data receiving frequency was examined, it was found that data was not coming in in a constant manner. This was due to the unstable wireless RSSI in the sensor room. This issue was resolved by changing to RS485 interface. Now data is coming in in a constant frequency. Below is the system block diagram.

Figure 4. System block diagram and modular hardware design.

The software application was developed with GUI in collaboration with National Instrument LABVIEW. This application will automatically detect the port and configure it. The only step involving a human resource is the keying in of the batch number. This creates identification (ID) for tracking purposes. Once the batch count is completed, the system will automatically reset the counter. If there is a malfunction of the system, a notification will be sent to the developer. After completion of each batch, detailed data will be sent to the developer via email. The data will be analyzed for improvement and also monitoring purposes. The report will be saved in a shared folder where the client will have access to.

Figure  5. Software application with GUI developed by LABVIEW

To conclude, even though we have achieved an accurate chicken count with reduced human resources, the issue of empty batch ID and out of alignment sensor still occurs. To fully eradicate this issue, industry internet (Industry 4.0) has to be implemented from the farm, suppliers and clients.